Because memory impairment is often the earliest effect of Alzheimer's disease (AD), dementia-related memory impairment must be detected and discriminated from aged related memory change. Long-term objectives are to distinguish memory changes in aging and early AD, detect preclinical and early AD, and validate criteria for diagnosis of preclinical AD. Specific aims are to: 1) improve prediction of clinical AD by detecting specific kinds of memory impairment in preclinical AD (i.e., encoding specificity) that are not secondary to decline of more basic cognitive abilities (e.g., processing speed); 2) distinguish memory and cognitive changes due to healthy aging from the memory and cognitive impairment due to preclinical AD; 3) test the hypothesis that unrecognized preclinical AD contributes to estimates of age-related cognitive decline in presumably non-demented adults; and 4) to determine if the mediators of cross-sectional age differences also mediate longitudinal cognitive decline in individuals. Our specific hypothesis are that: 1) identification of preclinical AD can be improved by detecting specific memory deficits in encoding specificity; 2) Preclinical AD will account for a significant amount of apparent age- related variance in cognitive measures; 3) Individual differences in processing speed and other basic mediators will account for cross- sectional age differences; 4) There will be a direct relation between age and encoding specificity that is not mediated by processing speed; 5) Mediators of age-related cognitive differences will not account for the cognitive differences related to preclinical AD; 6) Intra-individual age- related decline in basic cognitive abilities will predict intra-individual decline in memory, but 7) Significant longitudinal age-related decline in memory will remain after accounting for status and change in processing speed and other power mediators of cross-sectional age differences. Regression analysis and structural modeling will be used to explain memory changes in aging and AD by impairment of other cognitive processes in tests of memory, encoding specificity, processing speed, verbal ability, processing capacity, attention, and executive function. The results will provide information needed to detect preclinical AD and predict clinical AD, discriminate age-from AD-related cognitive changes, investigate cognitive aging without dementia, provide early treatment of AD, and correlate cognitive, physiologic, biochemical, neuropathologic, and neuroimaging changes in aging and AD.